Lunit INSIGHT MMG logo

Lunit INSIGHT MMG

AI-assisted breast cancer detection for 2D mammograms with malignancy scoring and lesion classification to support radiologist interpretation.

Solution by Lunit
Visit website

Overview

Lunit INSIGHT MMG is an AI-powered software solution designed to support radiologists in the detection of breast cancer from 2D mammograms. Trusted globally, it is built for high-volume breast imaging environments and is particularly suited to settings facing a shortage of breast imaging specialists. The platform delivers malignancy likelihood scores, classifies lesions, and highlights key areas of interest to enhance diagnostic confidence and streamline reading workflows.

Lunit INSIGHT MMG is designed for seamless integration with PACS and existing imaging viewers, enabling AI-driven insights to be delivered directly during image review. It supports both single- and double-reading workflows, helping radiology teams interpret images faster while reducing cognitive load and improving consistency in interpretation — ultimately contributing to better patient outcomes.

Core Capabilities

  • Malignancy likelihood scoring: Provides AI-generated scores indicating the likelihood that a lesion is cancerous, helping radiologists prioritise and interpret findings with greater confidence.
  • Lesion classification: Automatically classifies detected lesions to support more structured and consistent reporting.
  • Key area highlighting: Draws attention to regions of interest within the mammogram, reducing the time required to identify suspicious findings.
  • Dense breast support: Delivers 18% higher accuracy in extremely dense breasts and 8% higher accuracy across all breast density categories compared to radiologist-only reading.
  • Interval cancer detection: Supports identification of cancers that may be missed between screening rounds, with studies suggesting up to 40% of breast cancer cases could be diagnosed earlier with AI support.
  • Reduced false positives: Improves specificity to reduce unnecessary callbacks, lowering patient anxiety and follow-up burden on clinical teams.

Clinically Validated Performance

  • In a JAMA Oncology study, Lunit INSIGHT MMG outperformed other commercially available AI solutions, achieving the highest accuracy and optimal sensitivity–specificity balance in detecting malignant lesions.
  • A peer-reviewed study published in Breast Cancer Research demonstrated superior screening accuracy across all breast density categories, with particular gains in extremely dense breast tissue.
  • A retrospective analysis showed that Lunit INSIGHT MMG correctly flagged suspicious findings in a mammogram that was originally overlooked, with the patient not diagnosed until two years later — illustrating the tool's potential for earlier diagnosis.
  • Research published in Radiology (Nanaa et al., 2024) supports the finding that AI assistance can enable earlier diagnosis in a significant proportion of breast cancer cases.

Workflow and Reading Support

  • Shortens image reading time by surfacing the most relevant findings quickly and efficiently.
  • Reduces cognitive load for radiologists, particularly in high-volume screening programmes.
  • Supports both single-reader and double-reader mammography workflows.
  • Helps address workforce pressures in breast imaging by augmenting the capacity of existing radiology teams.
  • Improves clarity and consistency in interpretation, helping to reduce patient callbacks and associated anxiety.

AI Scoring — Key Clarifications

  • The malignancy likelihood score reflects the AI's confidence that a cancerous lesion is present — it does not indicate prognosis, severity, or treatment outcome.
  • The scoring scale is not linear; a higher score indicates greater AI confidence in malignancy, not a proportionally worse outcome.
  • The AI does not learn in real time from clinical images submitted during routine use.

Lunit INSIGHT MMG is designed for integration into existing PACS and imaging viewer environments, making deployment straightforward within established radiology workflows. The solution has been adopted in high-profile programmes including BreastScreen NSW in Australia and Sol Radiology in the United States. It is part of the broader Lunit INSIGHT Breast Suite, which also includes Lunit INSIGHT DBT for 3D tomosynthesis review. Availability may vary by country and region.

Meta

Domain
Digital Pathology & Imaging
Subdomain
AI Cancer Diagnostics
Software type(s)
Analytical Platform
Deployment type(s)
Cloud / SaaS
Industry vertical(s)
Academic / ResearchBiotechDiagnostics / IVDPharma
Development stage(s)
ClinicalPost-Market & RWE
Target user(s)
Research ScientistClinical / Diagnostic Professional
Compliance standard(s)
EU MDR
Tag(s)
Uses AI